More often than not, we’re in sync: patient and caregiver well-being over time in stem cell transplantation

Cancer impacts not only patients, but also their family members and informal caregivers. There are approximately 3.2 million unpaid family cancer caregivers in the United States today [1]. The number of caregivers needed to support cancer patients continues to rise [2]. As cancer treatment at large moves toward the outpatient setting, caregivers will be expected to provide increasing levels of unpaid care [3], potentially contributing to the already known physical and psychological morbidity associated with caregiving [4]. Caregiving for cancer patients can increase risk for caregivers’ social isolation, sleep problems [5], loneliness [6], depression, and anxiety [7].

The emotional toll of cancer caregiving can have detrimental effects on both caregivers and their patients. Indeed, a large body of evidence suggests that caregivers’ and their patients’ psychological functioning are inextricably linked [8, 9] such that when caregivers become more depressed, patients are likely to experience similar depressive symptoms, and vice versa. This reciprocity is often termed interdependence [10] or, in more stark terms, an emotional contagion [11]. Thus, there is strong empirical evidence that patient-caregiver dyads overlap in their emotional experience, supporting the theoretical basis for examining dyadic interdependence. While there is agreement that this interdependence occurs over the course of cancer treatment, we know much less about how and when interdependence within patient-caregiver dyads may change over time [12].

Psychological interdependence among patient-caregiver dyads is particularly important during hematopoietic stem cell transplant (HSCT). HSCT is an aggressive medical procedure in which patients’ bone marrow is ablated with chemotherapy, with or without radiation, donor cells are carefully matched to the patients and then infused into the patient, and the medical team supports the patient while the immune system is reconstituted over the next year or more [13]. For patients, emotional distress typically peaks leading into HSCT, and symptomatology continues to increase through the course of transplant during hospitalization. Patients’ anxiety and depression begin to decrease 3 and 6 months following HSCT [14]. For caregivers, distress is also high leading up to HSCT, with depression and anxiety increasing during the hospitalization and course of transplant [15]. Caregivers’ physical and emotion well-being can deteriorate in the months following transplant [16].

Despite the interdependent nature of patients and caregivers’ well-being during the HSCT process, relatively little work has examined this dynamic over time [12]. For the current study, we sought to address this gap by drawing from patients undergoing HSCT in the context of a stress management intervention targeting the caregiver. The intervention, called PEPRR (PsychoEducation, Paced Respiration, and Relaxation) [17], was effective at reducing caregiver distress [18], yet dyads’ interdependence over time was not investigated. Thus, our goal was to examine the potential interdependence of caregiver well-being and patient quality of life and, potentially, how the PEPPR intervention impacted dyad’s shared well-being during HSCT. In line with other dyadic research in HSCT populations [19] as well as our work with this study’s baseline data [20], we hypothesized that the emotional well-being of patients and caregivers would demonstrate a stronger interdependence over time in the PEPPR group, as caregivers may attend to their own emotional needs, thereby growing the emotional bond within the dyad.

These secondary data analyses drew from a 1:1 randomized clinical trial in which HSCT patients and their caregivers were randomly assigned to either caregivers receiving PEPPR or enhanced treatment as usual (eTAU; intervention materials were provided without prompting or coaching) [18]. The 4 study timepoints were anchored to patients’ day zero of HSCT (baseline) at which point questionnaire data was collected, as well as at 6 weeks, 3 months and 6 months post-transplant. Patient and caregivers completed the measures described below at these timepoints with baseline measures administered prior to randomization. The Colorado Multiple Institutional Review Board approved the study (www.​ClinicalTrials.​gov identifier: NCT02037568).

Allo‐HSCT patient/caregiver dyads (n = 407) were approached at the two participating hospitals and 331 dyads met eligibility criteria. One hundred and fifty-four patient and caregivers participated in this study and provided baseline data for analyses. As previously reported [18], there were no differences on baseline demographics between those assigned to PEPPR as compared to eTAU. Greater caregiver age was also significantly related to lower caregiver depression at baseline (r = − 0.22; p = 0.006) and younger patient age was related to worse patient QOL (r = 0.17; p = 0.031). Accordingly, all models accounted for baseline age.

Caregivers are critical to patients during the cancer journey, particularly in HSCT. There is emerging data highlighting that patient and caregiver well-being is often interdependent [20], but may change over time [37, 38]. The current study aimed to examine the interdependence of patients’ QOL and caregivers’ anxiety and depression. Overall, it is noteworthy that the data fit one parsimonious model as opposed to a multigroup SEM for the PEPPR and eTAU groups as originally hypothesized suggesting similar associations between dyads and across time in these two groups. In the full sample, we observed significant actor effects over the course of HSCT (patients’ QOL scores predicting their subsequent QOL and caregivers’ depression predicting their own subsequent depression). Additionally, we detected somewhat surprising partner effects, suggesting the interdependence changes over time; namely reverses direction such that if a patient has worse QOL at 6 weeks, their caregiver will be doing better (lower depression) at the 3-months. Finally, in visual comparison across the models examining caregivers’ anxiety and depression, there is some suggestion of greater interdependence for patient QOL to caregiver depression, than for patient QOL to caregiver anxiety across HSCT.

It is first worth commenting on the finding that the data did not fit a multigroup model as originally hypothesized. There are several reasons why the constrained model was retained over the multigroup model. We may have been underpowered to detect group differences with multiple estimated parameters, a common concern in longitudinal SEM’s [39]. Further, while the PEPPR intervention demonstrated a significant effect on caregiver distress [18], this effect may not be robust to extend to patient QOL, or dyads’ interdependence over time. As demonstrated in exploratory regression models, a number of medical factors likely impact QOL (length of hospitalization) and may overshadow any dyadic relationships of well-being. Future studies will benefit from more repeated measures of patients and caregivers, potentially harnessing technologies to track changes over more granular periods of time [40] and closely monitoring how changes in patient status relate interdependence.

The dyadic changes that we detected over the course of transplant, while small in effect, are intriguing for several reasons. First, the significant relationship that we expected between FC mood and patient QOL at baseline based on prior work [20] continued at 6 weeks but was no longer significant at 3 months. This may relate to patient and caregivers’ interrelatedness changing as care transitions to an outpatient setting in which more care is provided by the medical team and, potentially, relieve the caregiver or impose additional details of patient care that caregivers feel they need to observe. Second, we detected an unanticipated change in the direction of the relationship between patient QOL at 6 weeks and caregivers’ depression at 3-months in that caregivers were less depressed at 3 months when their patient had poorer QOL at 6-weeks post-HSCT. Changes in patients’ medical treatment—experienced by both members of the dyad—may also explain changing interdependence. We attempted to explore this hypothesis by looking at the impact of hospitalized days as predictors of in the regressions suggesting that the greater number of days that patients were in the hospital approached significance as a predictor of greater depression at 3-months post-transplant. Future research can build on these initial findings by closely monitoring changes in patients’ medical status.

The primary analyses with the current data showed the intervention improved distress overall, with slightly greater effect sizes for anxiety than depression. The current analyses explored dyadic effects within these constructs and although the associations among patient quality of life and caregiver mental health did not differ between the intervention and control conditions, dyadic interdependence may still be important to consider when designing and evaluating interventions. This work provides evidence of dyadic interdependence which may affect the success, or lack thereof, of an intervention. For instance, there is evidence of strong mutual influences on health between close relationship partners [41]. Physical health problems often negatively impact mental health of partners in close relationships; at the same time, interdependence can be leveraged to improve intervention effectiveness [41]. In addressing health behavior, interventions targeting partners can help patients make behavior change, by reinforcing positive behaviors [42]. This model has been extended to how shared health behaviors and emotional experience can impact biobehavioral pathways that impact health and disease [43] and our group has demonstrated how the shared behavior withing couples (e.g., sleep quality) can impact immune reconstitution in HSCT [44]. Overall, dyadic interdependence remains an important consideration in designing and testing interventions, particularly in HSCT.

The current findings should be qualified with several limitations. First, the patients and caregivers recruited were fairly homogenous, with the majority of caregivers being white females and patient-caregiver dyads being spousal. Future research should seek a more diverse demographic sampling to increase generalizability and examine whether similar patterns of interdependence are observed in other patient-caregiver relationships (e.g., parent–child dyads). Additionally, the current analyses represent secondary data analyses and the original study may not have been powered to detect dyadic relationships over time, as multigroup SEMs often require hundreds of participants [39]. Larger samples will allow for additional mediation or moderation models to further explore dyadic relationships [45], or potentially, different mechanisms [46] of intervention effects. Finally, we had a relatively high rate of dropout at the later timepoints in this study (> 40% at the 6-month timepoint). While we attempted to address the assumptions of data being missing at random [34] and our findings did not suggest significant differences between those providing data at the final 6 months timepoint and those lost to follow-up, it is possible that our findings would only generalize to a population of patients and caregivers that are willing and able to sustain participation in a longitudinal study.

This study highlights the interrelated, yet nuanced, patient and caregiver relationship during the course of stem cell transplantation. These results highlight the need for not only additional support for the patient, but also the caregiver and possibly the closely connected dyad. While future work may identify how these change in the context of the intervention, our findings imply that the association between patient QOL and partner depression may differ across the course of transplant and interventions that seek to improve patient-caregiver well-being should remain mindful of these differences when study dyadic relationships. Future research should explore these relationships and include longer-term follow-up to identify when, and how, to better support patients and their closely connected caregivers.